USING A COMPLEX ADAPTIVE SYSTEMS APPROACH TO UNDERSTAND ANTIMICROBIAL RESISTANCE

Background: Complex adaptive systems are dynamic open systems made up of individual agents that act independently but whose actions are interconnected so that when one agent changes it has the potential to impact other agents and create a ripple effect across the system. Complex adaptive systems abound in public health and antimicrobial resistance (AMR) is an example of one. AMR is a growing issue that poses a significant health and economic burden worldwide. A better understanding of all the factors influencing AMR will better identify actions most likely to succeed in reducing antimicrobial use and the burden posed by resistant infections. Using complexity science, the main goal of this study was to create a visual model of AMR in Canada, showing the different parts of the ‘system’, and reflecting the types of individuals and factors playing a role. Methods: We conducted four face-to-face meetings and several key informant interviews to gather input for the model. In order to build on and advance current models of AMR, we invited people who represented diverse viewpoints not traditionally considered when discussing the issue of AMR to participate in the study. For each meeting, maximum variation sampling was used to select a group of potential participants representing different sectors and disciplines, age groups, genders, and geographic locations across Canada. We developed an initial model represented as a causal loop diagram to serve as a starting point for discussions. The development of the model was intentionally iterative with input from participants at each meeting and interview used to revise the model. Results: This ongoing study continues until March 2017. A wide range of individuals have participated to date, including: an economist, a food security practitioner, a dentist, a pharmacist, a swine producer, a food retail representative and a farm animal welfare representative. Despite participants having a diverse range of backgrounds and areas of expertise, the discussions were flowing, interactive and congenial. The data collected to date have greatly expanded and transformed the initial model; new variables were suggested for areas throughout the entire model and a number of variable name changes were suggested for clarity, greater inclusivity, and better representation. Conclusions: The end result of this study will be a qualitative, conceptual model that will visually describe the complex nature of the AMR system in Canada. Derivative maps will be developed to highlight particular variables and pathways, and help provide clarity to the full visual model. Assessment of the model will enable the identification of critical leverage points which are those aspects of the system that need to shift or change so that the whole system can adapt. It is envisioned that at the end of this study, the individuals, agencies and organizations described in this conceptual model will recognize their shared responsibility in the AMR system and their role in shifting or enabling the model to adapt to a new reality.